You are on page 1of 7

The Journal of Maternal-Fetal and Neonatal Medicine, October 2006; 19(10): 633638

Differences in amniotic fluid patterns and fetal biometric parameters in


third trimester pregnancies with and without diabetes

ALEXANDER KOFINAS & GEORGE KOFINAS


Department of Obstetrics and Gynecology, Kofinas Perinatal and Fertility Institute, New York Methodist Hospital,
Brooklyn, New York
(Received 10 March 2006; revised 30 April 2006; accepted 8 May 2006)

Abstract
Objective. The amniotic fluid index (AFI) has been increasingly used in the assessment of fetal well-being. We conducted
the study to evaluate and compare the amniotic fluid index in third trimester normal and diabetic human pregnancy and to
assess the correlation between the AFI and the fetal biometric parameters.
Methods. Real-time ultrasound was performed to evaluate the AFI (four-quadrant technique), and to measure the biparietal
diameter, head circumference, abdominal circumference, and femur length in 225 normal and 120 diabetic pregnant women
from 27 to 42 weeks of gestation. Each patient was studied only once.
Results. AFI in normal pregnancies was less than that in diabetic pregnancies throughout the gestational ages studied
(2742 weeks). In normal pregnancy, the mean AFI was 14.0 cm at 27 weeks and decreased to 11.4 cm at 42 weeks
(r 0.25, p 0.0005), whereas in diabetic pregnancies, the values remained stable throughout the gestational ages studied.
There exist significant differences in AFI, estimated fetal weight, estimated fetal weight %, abdominal circumference,
abdominal circumference %, and head circumference to abdominal circumference ratio between the two groups. In both
normal and diabetic pregnancies, there is a positive correlation between the AFI and the percentile of abdominal
circumference (p 5 0.0001), and between the AFI and the percentile of estimated fetal weight (p 5 0.0001).
Conclusion. This study provides gestational age-specific values of the AFI in normal and diabetic pregnancies. Diabetic
pregnancy has greater AFI values than normal pregnancy between 27 and 42 weeks. The AFI correlates to the percentile of
the estimated fetal weight and the abdominal circumference in both groups, suggesting that there may be a relationship
between increased AFI and large for gestational age fetus independent of diabetes.

Keywords: Pregnancy, amniotic fluid, diabetes, biometry

Introduction
Estimation of amniotic fluid volume is an important
part of fetal assessment, especially during the third
trimester. Assessment of amniotic fluid volume by
clinical means alone is difficult and inaccurate, but
real-time sonography simplifies the evaluation and
provides accurate amniotic fluid volume estimation.
Phelan and colleagues were first to describe the
amniotic fluid index (AFI) [1,2]. Subsequently,
several investigators have reported values of the
AFI for specific weeks of gestation [37].
The relationship between abnormal AFI and
unfavorable perinatal outcome has been reported.
Fetal conditions that are associated with oligohydramnios include intrauterine growth restriction,
post maturity, and major fetal anomalies, commonly

involving the fetal genitourinary tract [811]. Fetal


conditions that are associated with hydramnios
include maternal diabetes mellitus, major congenital
anomalies such as open neural tube defect, upper
gastrointestinal obstruction or malformations, congenital infections, and both immune and nonimmune forms of hydrops fetalis [1114]. For both
oligohydramnios and hydramnios, there is increased
perinatal morbidity and mortality. A significant
number of pregnancies have evidence of hydramnios
in the absence of any apparent pregnancy complications. The relationship between fetal size and
amniotic fluid volume has been suggested [12,14].
The objective of this study was to evaluate the
amniotic fluid index in the third trimester of normal
and diabetic pregnancies and to evaluate the relationship between the AFI and fetal biometric parameters.

Correspondence: Alexander Kofinas, MD, 901 Stewart Av., Suite 245, Garden City, NY 11530, USA. Tel: 1 516 832 0300. Fax: 1 516 832 0301.
E-mail: unborn@kofinasperinatal.org
ISSN 1476-7058 print/ISSN 1476-4954 online 2006 Informa UK Ltd.
DOI: 10.1080/14767050600822547

634

A. Kofinas & G. Kofinas

Methods
We evaluated 225 normal pregnancies and 120
pregnancies complicated by diabetes at 27 to 42
weeks of gestation. All patients were referred to our
unit for routine normal or diabetic prenatal ultrasound evaluation, and ultrasonic estimation of fetal
weight between January 2001 and December 2004.
All normal pregnant women had uncomplicated
singleton pregnancies. Dating was established by
accurate menstrual history confirmed by sonography
prior to 20 weeks. Pregnancies complicated by
diabetes consisted of patients with gestational
(n 95) or uncomplicated pregestational diabetes
(n 25). Patients with vasculopathy, renal disease,
intrauterine growth restriction (IUGR), chronic
hypertension and/or preeclampsia, were excluded.
All patients with diabetes were delivered at no later
than 40 weeks of gestation. The study design was
cross-sectional and only the first sonographic evaluation from each patient was used. The study was
approved by the institutional review board.
Ultrasound scans were performed with a real-time
scanner with a 3.75 MHz curved-linear transducer.
The amniotic fluid index was obtained by the fourquadrant technique described by Phelan et al. [1,2].
For each patient studied, the fetal biometric
parameters measured were biparietal diameter, head
circumference, abdominal circumference, and femur
length. The estimated fetal weight, fetal weight
percentile, and the various biometric ratios were
assessed by means of computerized analysis based on
previously reported fetal biometric studies [15].
Statistical analysis was performed by means of
JMP Statistical Discovery Software for personal
computers (SAS Institute Inc., Cary, NC, USA).

The means of the AFI, abdominal circumference,


abdominal circumference percentile, estimated
fetal weight, fetal weight percentile, and head
circumferenceabdominal circumference ratio of
the two groups were compared by t-test. Pearsons
correlation analysis was used to evaluate the correlation between the AFI and the various biometric
parameters.
Results
There is a significant inverse correlation; the AFI
declines from an average of 14.0 cm at 27 weeks to
11.0 cm at 42 weeks (p 5 0.002). There is no
correlation between the AFI and gestational age;
the AFI value remains stable from 27 to 40 weeks of
gestational age. The mean AFI in normal pregnancies was less than the mean AFI in pregnancies
complicated by diabetes throughout the gestational
ages studied, and the difference became more
prominent with advancing gestational age (Figure 1).
Table I presents the comparison of the means of
the AFI and the biometric parameters between
normal pregnancies and pregnancies complicated
by diabetes. AFI in normal pregnancies was
13.2 + 0.3 cm in comparison to 14.6 + 0.4 cm in
patients with diabetes (p 5 0.002). As noted,
significant differences exist between the two
groups when the various biometric parameters are
compared.
Figure 2 depicts the linear fitting of the abdominal
circumference percentile against the AFI in normal
pregnancies and Figure 3 in pregnancies complicated
by diabetes. AFI increases with increasing values of
abdominal circumference percentile (p 0.0001).
Figure 4 shows a significant positive linear relationship

Figure 1. Concomitant plotting of AFI in normal pregnancies and in pregnancies with diabetes against gestational age. Note the different
slopes.

Amniotic fluid patterns in pregnancies with diabetes


between the estimated fetal weight percentile and
the AFI in normal pregnancies and Figure 5 in
pregnancies complicated by diabetes (p 0.0001).
Discussion
The amniotic fluid index provides a semi quantitative
analysis of the amniotic fluid volume. The technique
is simple and highly reproducible. The intra-observer
and inter-observer variations have been found to be
small [3,16,17], and can relate to fetal movement
[18].
This study provides gestational age-specific values
of the AFI in normal pregnancies in the third
trimester, which are in agreement with previous
reports by Phelan et al. [2], Moore et al. [3], Jeng
et al. [4], and Hallak et al. [5]. To our knowledge,

Table I. Comparison of fetal biometric parameters in normal and


diabetic pregnancies.

Parameter
AFI
EFW*
EFW %
AC
AC %
HC/AC**

Diabetes
Mean (SE)
14.6 cm
2394.9 g
63.2
30.2 cm
65.8
1.05

(0.4)
(77.0)
(2.4)
(0.3)
(2.2)
(0.006)

Normal
Mean (SE)
13.2 cm
2063.3 g
51.8
28.3 cm
54.4
1.09

(0.27)
(56.0)
(1.8)
(0.3)
(1.6)
(0.005)

p Value
0.002
0.0006
0.0002
0.0001
0.0001
0.0001

SE, standard error; AFI, amniotic fluid index; EFW, estimated


fetal weight; AC, abdominal circumference; HC, head circumference. *Estimated fetal weight at time of ultrasound. **Head
circumference to abdominal circumference ratio.

635

this is the first report on the AFI curve of diabetic


pregnancies during the 3rd trimester. As it becomes
evident from the correlation analysis, the AFI of
diabetic patients follows a different pattern than the
AFI of normal patients. This finding may have
significant implications in the clinical application of
AFI measurements in fetal well-being assessment. It
is important to note here that all diabetic patients
were delivered at no later than 40 weeks of gestation
and thus we cannot tell whether the AFI would
remain stable or decline between 40 and 42 weeks.
The production and regulation of amniotic fluid is
a dynamic and complex process involving mainly
fetal urine output, fetal swallowing, and fetal lung
fluid flow [19]. Chamberlain et al. [12], and Varma
et al. [14] described the possibility that amniotic
fluid volume is related to fetal weight, but the
mechanism of this observation is not understood.
Three different theories [20] have been introduced to
explain the possible interaction between maternal
glycemic status and amniotic fluid volume: (1)
maternal hyperglycemia induces fetal hyperglycemia
resulting in osmotic diuresis when the fetal threshold
for glucose is exceeded; (2) as glucose equilibrates
across the placenta there is an isosmotic movement
of fluid toward the fetal compartment with volume
expansion and an increase in glomerular filtration
rate leading to enhanced fetal urine output production; and (3) decreased fetal swallowing without
concomitant change in fetal urination.
It has been speculated that hydramnios noted in
pregnancies with diabetes is the result of osmotic
diuresis secondary to fetal hyperglycemia and

Figure 2. AFI plotted against abdominal circumference percentile in normal pregnancies. There is significant positive correlation.

636

A. Kofinas & G. Kofinas

Figure 3. AFI plotted against abdominal circumference percentile in pregnancies with diabetes. There is significant positive correlation.

Figure 4. AFI plotted against estimated fetal weight percentile in normal pregnancies. Note the significant positive correlation.

glycosuria. However, the concentration of glucose


and other solutes in amniotic fluid is not consistently
related to amniotic fluid volume [21,22]. Fetuses of
diabetic pregnancies spend more time breathing than
fetuses of normal pregnancies [23]. We speculate
that since by nature breathing and swallowing are

mutually exclusive, increased time devoted to fetal


breathing reduces the time that the fetus spends
swallowing. Decreased swallowing activity may
decrease in part the amount of amniotic fluid
removed from the intra-amniotic space, which in
turn causes increased amniotic fluid volume.

Amniotic fluid patterns in pregnancies with diabetes

637

Figure 5. AFI plotted against estimated fetal weight percentile in pregnancies with diabetes. Note the significant positive correlation.

Diabetic patients have greater AFI, abdominal


circumference, and estimated fetal weight than
normal pregnant patients. As expected, the head
circumferenceabdominal circumference ratio is
smaller in the diabetic group indicative of the
diabetic type of asymmetrical growth.
Our findings confirm the higher incidence of
increased amniotic fluid volume (hydramnios) and
large for gestational age (macrosomia) in the diabetic
patients. Interestingly, the AFI correlates to the
percentile of the estimated fetal weight and the
abdominal circumference in both diabetic and
normal pregnancies. This suggests that there is a
relationship between increased AFI and large fetal
size during the third trimester for both diabetic and
normal pregnancies. This relationship was identified
in normal pregnancies before by means of AFI and
deepest vertical amniotic fluid pocket [24,25]. It is of
interest to note that regardless of the presence of
diabetes, fetal size normalized for gestational age is in
part responsible for the higher amniotic fluid volume
in such patients.
In summary, this study suggests a correlation
between the amniotic fluid volume (expressed as
AFI) and the estimated fetal weight and abdominal
circumference corrected for gestational age.
Although, fetal weight variability alone cannot
explain the presence of hydramnios in the apparently
normal fetus, we suggest that this information is
taken into account when patients are counseled for
the potential risks of hydramnios.

References
1. Phelan JP, Smith CV, Broussard P, Small M. Amniotic fluid
volume assessment using the four-quadrant technique at
3642 weeks gestation. J Reprod Med 1987;32:540542.
2. Phelan JP, Ahn MO, Smith CV, Rutherford SE, Anderson E.
Amniotic fluid index measurements during pregnancy.
J Reprod Med 1987;32:601604.
3. Moore TR, Cayle JE. The amniotic index in normal human
pregnancy. Am J Obstet Gynecol 1990;162:11681173.
4. Jeng CJ, Jou TJ, Wang KG, Yang YC, Lee YN, Lan CC.
Amniotic fluid index measurement with the four-quadrant
technique during pregnancy. J Reprod Med 1990;35:674
677.
5. Hallak M, Kirshon B, Smith EO, Cotton DB. Amniotic fluid
index. Gestational age-specific values for normal human
pregnancy. J Reprod Med 1993;38:853856.
6. Sepulveda W, Flack NJ, Fisk NM. Direct volume measurement at midtrimester amnioinfusion in relation to ultrasonographic indexes of amniotic fluid volume. Am J Obstet
Gynecol 1994;170:11601163.
7. Nwosu EC, Welch CR, Manasse PR, Walkinshaw SA.
Longitudinal assessment of amniotic fluid index. Br J Obstet
Gynaecol 1993;100:816819.
8. Chamberlain PF, Manning FA, Morrison I, Harman CR,
Lange IR. Ultrasound evaluation of amniotic fluid volume. I.
The relationship of marginal and decreased amniotic fluid
volumes to perinatal outcome. Am J Obstet Gynecol 1984;
150:245249.
9. Manning FA, Hill LM, Platt LD. Qualitative amniotic fluid
volume determination by ultrasound: Antepartum detection
of intrauterine growth retardation. Am J Obstet Gynecol
1981;139:254258.
10. Sarno AP, Ahn MO, Phelan JP. Intrapartum amniotic fluid
volume at term: Association of ruptured membranes, oligohydramnios and increased fetal risk. J Reprod Med 1990;35:
719723.

638

A. Kofinas & G. Kofinas

11. Anandakumar C, Biswas A, Arulkumaran S, Wong YC,


Malarvishy G, Ratnam SS. Should assessment of amniotic
fluid volume form an integral part of antenatal fetal
surveillance of high risk pregnancy? Aust NZ J Obstet
Gynaecol 1993;33:272275.
12. Chamberlain PF, Manning FA, Morrison I, Harman CR,
Lange IR. Ultrasound evaluation of amniotic fluid volume. II.
The relationship of increased amniotic fluid volume to
perinatal outcome. Am J Obstet Gynecol 1984;150:250254.
13. Benson CB, Coughlin BF, Doubilet PM. Amniotic fluid
volume in large-for-gestational-age fetuses of nondiabetic
mothers. J Ultrasound Med 1991;10:149151.
14. Varma TR, Bateman S, Patel RH, Chamberlain GV, Pillai U.
The relationship of increased amniotic fluid volume to
perinatal outcome. Int J Gynecol Obstet 1988;27:327333.
15. Hadlock FP, Harrist RB, Sharman RS, Deter RL, Park SK.
Estimation of fetal weight with the use of head, body, and
femur measurementsa prospective study. Am J Obstet
Gynecol 1985;151:333337.
16. Rutherford SE, Smith CV, Phelan JP, Kawakami K, Ahn MO.
Four-quadrant assessment of amniotic fluid volume. Interobserver and intraobserver variation. J Reprod Med
1987;32:587589.
17. Peedicayil A, Mathai M, Regi A, Aseelan L, Rekha K,
Jasper P. Inter- and intra-observer variation in the amniotic
fluid index. Obstet Gynecol 1994;84:848851.

18. Wax JR, Costigan K, Callan NA, Gegor C, Johnson TR.


Effect of fetal movement on the amniotic fluid index. Am J
Obstet Gynecol 1993;168:188189.
19. Gilbert WM, Moore TR, Brace RA. Amniotic fluid volume
dynamics. Fetal Med Rev 1991;3:89104.
20. Moore TR. Diabetes in pregnancy. In: Creasy RK, Resnik R,
editors. Maternal fetal medicine, principles and practice.
Philadelphia: WB Saunders; 1994. p 934.
21. Van Otterlo L, Wladimiroff J, Wallenburg H. Relationship
between fetal urine production and amniotic fluid volume in
normal pregnancy and pregnancy complicated by diabetes.
Br J Obstet Gynaecol 1977;84:205209.
22. Dashe JS, Nathan L, McIntire DD, Leveno KJ. Correlation
between amniotic fluid glucose concentration and amniotic
fluid volume in pregnancy complicated by diabetes. Am J
Obstet Gynecol 2000;182:901904.
23. Devoe LD, Youssef AA, Castillo RA, Croom CS. Fetal
biophysical activities in third-trimester pregnancies complicated by diabetes mellitus. Am J Obstet Gynecol 1994;
171:298303.
24. Myles TD, Nguyen TM. Relationship between normal
amniotic fluid index and birth weight in term patients
presenting for labor. J Reprod Med 2001;46:685690.
25. Myles TD, Morgan JL, Santolaya-Forgas J. Deepest vertical
amniotic fluid pocket at term. Normal values and clinical
application. J Reprod Med 2003;48:712.